Personalized Visualization Recommendation System

2. The method of claim 1, further comprising: generating first information and second information based on monitoring the user interactions, wherein the first information relates a plurality of users with the plurality of data attributes and the second information relates the plurality of users with the plurality of visualizations; and generating third information by computing aggregate statistical properties of data corresponding to each of the data attributes, wherein the third information comprises the plurality of meta-features.

3. The method of claim 1, wherein: the model is generated based on the user interactions with the plurality of visualizations of the plurality of datasets and the plurality of meta-features representing the plurality of data attributes for each of the plurality of datasets, and wherein the user interactions include at least one from a set comprising clicking on elements of the visualizations, hovering on elements of the visualizations, selection of chart types for the visualizations, utilization of data transformations for the plurality of datasets, and selection of colors for the visualizations.

4. The method of claim 2, further comprising: performing joint matrix factorization on the first information, the second information, and the third information to learn the low-dimensional embeddings.

5. The method of claim 2, further comprising: performing joint tensor factorization on the first information, the second information, and the third information to learn the low-dimensional embeddings.

6. The method of claim 1, further comprising: generating a list of recommended visualizations for the user based on the predicted visualization preference weights; and receiving a selection input from the user for selecting one of the recommended visualizations, wherein the personalized visualization recommendation is based on the selection input.

7. The method of claim 1, further comprising: displaying data from at least one of the plurality of datasets based on the personalized visualization recommendation.

8. The method of claim 1, further comprising: performing unsupervised data clustering to generate one or more landmark features, wherein the plurality of meta-features includes the one or more landmark features.

9. The method of claim 1, further comprising: identifying an additional dataset comprising additional data attributes; updating the model based on the additional data attributes; and generating an additional personalized visualization recommendation of the additional dataset based on the updated model.

10. The method of claim 1, further comprising: identifying an additional user; updating the model based on the additional user; and generating an additional personalized visualization recommendation for the additional user based on the updated model.

11. The method of claim 1, further comprising: identifying additional user interactions; updating the model based on the additional user interactions; and updating the personalized visualization recommendation based on the updated model.

12. The method of claim 1, further comprising: identifying a pre-determined number of data attributes; and selecting a subset of the data attributes based on the pre-determined number, wherein the personalized visualization recommendation is based on the selected subset of data attributes.

13. A method for data visualization, comprising: identifying a first matrix that relates a plurality of users with a plurality of data attributes of a plurality of datasets; identifying a second matrix representing user interactions with a plurality of visualizations; identifying a third matrix representing a plurality of meta-features for each of the data attributes; performing joint factorization of the first matrix, the second matrix, and the third matrix to obtain a plurality of joint factorization matrices, wherein a first joint factorization matrix of the plurality of joint factorization matrices represents a first factor of the first matrix and a factor of the second matrix, a second joint factorization matrix of the plurality of joint factorization matrices represents a second factor of the first matrix and a factor of the third matrix; generating a model for predicting visualization preference weights based on the plurality of joint factorization matrices; predicting the visualization preference weights for a user corresponding to a plurality of candidate visualizations of at least one dataset using the model; and generating a personalized visualization (or visualization recommendation) for the at least one dataset for the user based on the predicted visualization preference weights.

14. The method of claim 13, further comprising: computing low-dimensional embeddings representing user characteristics, the data attributes, visualization configurations, and the meta-features using joint factorization of the first matrix, the second matrix and the third matrix, wherein the low-dimensional embeddings comprise a fourth matrix representing the user characteristics, a fifth matrix representing the data attributes, a sixth matrix representing the visualization configurations, and a seventh matrix representing the meta-features.

15. The method of claim 14, further comprising: performing a coordinate descent operation, wherein the low-dimensional embeddings are based on the coordinate descent operation.

16. The method of claim 14, further comprising: identifying at least one regularization term, wherein the low-dimensional embeddings are computed based on the at least one regularization term.

17. The method of claim 14, further comprising: receiving an additional dataset comprising additional data attributes; updating the first matrix and the third matrix based on the additional data attributes; and updating the low-dimensional embeddings based on the updated first matrix and the updated third matrix.

19. The apparatus of claim 18, wherein: the data collection component is configured to monitor the user interactions with the plurality of data attributes for each of the plurality of datasets and the user interactions with the plurality of visualizations of the plurality of datasets.